ProtoLabs 3D Printing: Six Pre-Order Checks After $14,000 in Mistakes

I owe this checklist to a box of useless plastic parts. Back in 2019, I ordered 18 small mounting brackets from a rapid prototyping service. Every bracket arrived at roughly one-twenty-fifth scale. My model was in inches; the upload form assumed millimeters. An STL file does not store units, so the system did the reasonable thing with the numbers it received. That mistake cost about $700 in material and a week and a half of schedule.

I am a mechanical engineer, and I have handled our company's outsourced manufacturing orders for five years. In that time I have logged roughly $14,000 in scrap and rework. Maybe a bit less; I would have to count. The cause was never a bad vendor. It was bad file prep, careless material selection, and the habit of rushing the last step. So I built a checklist, and my team now runs it before every 3D printing order we send to ProtoLabs. It has caught more near-misses than I care to admit.

Here are the six checks, in the order of how often each one has bitten us.

Check 1: Do you actually need to buy a printer? Read this before any 3D printer review

If you are searching for things like “ProtoLabs industrial 3D printer reviews” or “ProtoLabs 3D printers,” you might be expecting to find a machine. ProtoLabs does not sell printers. It runs a digital manufacturing service that includes industrial 3D printing. The real question is not which machine to review. It is whether buying a machine beats using a service for your actual workload.

A year ago, I almost convinced my boss to buy an SLS machine. I spent two weekends reading industrial 3D printer reviews and comparing build volumes. I had a spreadsheet that said we would save $18,000 per year by printing in-house.

The spreadsheet was wrong. It used the cheapest material price and pretended that maintenance, calibration, failed builds, and training cost nothing. Our real utilization would have been around three or four hours a week. An expensive machine that sits idle is not an asset.

That old assumption—that a serious product team should own its 3D printers—comes from an era when outsourcing meant emailing an unknown shop and waiting days for a quote. That world has changed. Digital manufacturing services now provide instant quotes, design analysis, and short lead times. Buying a printer can still be the right call if you print constantly or need tight iteration loops. But do not decide from review sites. Decide from your order history.

If you are evaluating a printer purchase, answer this number before anything else: how many hours per week will it actually run?

A related question I hear often: do 3D printers require a computer? In my experience, yes. You need to create and slice the model on a computer. Some machines will print from an SD card once the file is sliced, and some have built-in touchscreens, but the file still has to be prepared somewhere. Software, operator time, and the learning curve are all part of the real cost of a printer. With a service like ProtoLabs, the only computer you need is the one you upload the file from.

Check 2: Check your units and the upload preview

The bracket mistake at the top of this post happened because I trusted a units menu and never looked at the preview. The CAD file looked perfect. The export dialog looked normal. An STL file has no scale, only coordinates.

Set your units deliberately before exporting, and then measure a known dimension in the viewer after upload. If your part should be 62 mm long but the preview reads something else, now is the time to catch it. Catching it after the parts arrive is how people end up owning 18 miniature brackets.

Check 3: Check wall thickness against the material rules

Thin walls look perfectly fine on screen. A 0.5 mm rib in CAD feels crisp and intentional. To an SLS machine, it is a gamble.

I once ordered a PA12 part with a few walls that were below the material's recommended minimum. The printer produced them, sort of. The part looked complete until the second time I flexed it, and then one wall snapped clean.

Before you upload, find the thinnest continuous section of your part and compare it with the design guidance for the material and technology you are choosing. ProtoLabs publishes design guides for this, and the quote process often includes DFM warnings. Treat them as constraints, not suggestions. If the wall does not need to be thin, make it thicker. It costs nothing in the digital file and it removes a failure mode from the physical part.

Check 4: Decide which face is allowed to look bad

Every 3D printing process leaves traces—layer lines, support marks, a slightly rough texture. If you never specify which surfaces matter, you will get a technically correct part with randomly placed cosmetics.

Open the model and think about orientation before you order. Which face is visible in the final assembly? Which hole has to stay round? What should supports never touch? If the upload interface lets you set orientation, set it. If you are unsure, write an order note: “The front face of this part is critical—please avoid supports there.” One sentence has saved us from more than one $400 round of parts that were functionally fine but embarrassing in a customer demo.

Check 5: Pick the material for your real conditions

A clear SLA resin looked like the right material for a prototype lens cover. It was transparent, it printed quickly, and the price was friendly. Nobody looked at the heat deflection temperature until after the lens, mounted 20 mm above a warm motor, began to sag during a demo. That was a fun meeting.

Data sheets list properties at specific conditions. Your part does not exist at “standard conditions”; it exists where you put it. Before choosing a material, write down three or four requirements: operating temperature, exposure to moisture, chemicals, or UV, how much load it carries, and whether it needs to survive more than one assembly cycle. Then compare the data sheet at those conditions. A one-off cosmetic check can use the cheap material. A functional test needs a material that removes a risk, not one that just looks right in the selector.

Check 6: Read the DFM feedback before you pay

ProtoLabs' speed comes from automation: upload a file, get a quote, get design-for-manufacturing feedback. That feedback is easy to skip when you are in a hurry. I skipped it in Q1 2024 on a fourteen-part order. The system flagged a boss that was near the minimum recommended geometry. I clicked past the warning because the deadline was tight.

The parts arrived, the boss was weak, and a fastener torqued it to failure during assembly. Three days lost on rework, all because I glanced at the bottom line instead of reading the note.

Read every warning line like it is the reason you are ordering. And if your part has a dimension that truly matters, say so. A comment like “the 4.2 mm bore on the left side is a press fit—please prioritize it” tells the manufacturing engineer where to put their attention. It beats hoping they guess.

Things this checklist will not fix

A 3D printed prototype is not the same thing as a production part, even when the material name matches. Printed nylon and molded nylon have different internal structure. Treat your first printed batch as a learning sample, not as proof that the design is done.

Also, do not copy an entire production drawing onto a prototype. Mark the two or three dimensions that actually affect fit, and know what the chosen process can hold. Tight tolerances on cosmetic features add cost and inspection time, and they make a prototype seem more final than it is.

And you will still make mistakes. I did last month—a material assumption that I should have verified. The difference is that now the checklist catches most problems while the part is still digital, when fixing it costs nothing. The best part of finally systematizing this process is that I no longer stay up the night before an order wondering if I missed a dimension. Run these six checks before your next ProtoLabs order. You will probably find something to correct. That is exactly the point.